Rapamycin is the trivial name of an extraordinary natural product produced by a soil micro-organism, and it was first discovered through its ability to halt fungal growth. It soon became clear that antifungal activity was only one activity of rapamycin. It has been most thoroughly investigated as an immunosuppressive agent and is undergoing clinical trials as an anticancer agent. Most small molecules bind to and inhibit a single large molecule--HIV protease inhibitors inhibit the enzyme HIV protease, aspirin and ibuprofen (the active ingredient in Advil and Nuprin) block cyclooxygenase, and so on. Rapamycin controls biological processes in a different way by binding two large molecules together. About five years ago we became involved in studying the atomic details of how rapamycin works by studying the structure of rapamycin bound to one of its targets, and a year ago we completed the picture by using X-ray diffraction to define the structure of rapamycin simultaneously binding two proteins together. These studies not only showed how rapamycin works, they also showed the atomic structures of the two previously undefined proteins that rapamycin binds to. Several investigators, particularly Stuart Schreiber at Harvard and Jerry Crabtree at Stanford, soon realized that rapamycin's ability to bind two proteins together was the key to controlling a wide variety of biological processes. To understand the implications of a small molecule binding two large molecules together, a bit of background is needed. Many biological processes are controlled by having one molecule bring two other together. For example, human growth hormone brings two human growth hormone receptors together, and the signal for a growth starts when the two receptors are brought together. What rapamycin does is the small molecule analog of this joining process. Rapamycin has a molecular weight of around 1000 daltons and can pass through cell membranes. Growth hormone, for example, has a molecular weight of 22,000 daltons and cannot pass through cell membranes. Together with our collaborators we're designing and studying a series of rapamcyin analogs as well as analogs of the two proteins they bind together. We envision that these constructs will be useful in a variety of ways, and perhaps the most important will be gene therapy. Problems in gene therapy can be broken down into two main issues: introducing the genes needed and controlling the introduced genes. How can genes be controlled? A typical method is through the use of 'transcription factors' or proteins that tell a particular stretch of DNA that its instructions should be transcribed into messenger RNA. Transcription factors have two domains called the DNA binding domain and the activation domain. Both the binding and activation domain must be engaged with DNA for the gene to be expressed; if only one domain is engaged, little happens. Rapamycin and rapamycin analogs have been used to bring together binding and activation domains and thereby control gene expression. Our colleagues at ARIAD Pharmaceuticals have engineered mice in which the gene for human growth hormone has been introduced and controlled with rapamycin. Rapamycin turns the gene on, and withdrawal of rapamcyin shuts the gene down.